Résumé :
|
[BDSP. Notice produite par INIST-CNRS 9R0x9u4D. Diffusion soumise à autorisation]. Surveillance is a critical public health tool for the control of pesticide poisoning. However, surveillance activities in developing countries are bedevilled by multiple problems, and inferences made from review of flawed data may lead to mistaken policy decisions. Results of intensified surveillance from an intervention project in the Western Cape Province of South Africa were compared to the pattern of poisonings reported in routine notifications to the health authorities for a control farming district and in the study district over a 5-year period preceding the study. Intensified surveillance data results were also contrasted with policy approaches based on routine notifications and on Regional Poison Centre reports. Poisoning rates reported in the study area increased almost 10-fold during the intervention period. Compared to intensified surveillance, hospital and health authority sources greatly underestimate the proportion of cases due to occupational poisoning, and overestimate suicide as a proportional cause. In addition, the risks for women appear underestimated from routine notifications. Assumptions that a lack of awareness is responsible for most poisonings are not borne out by the empirical data when reporting is intensified. Current policy assumptions are faulty, may result in inappropriate blame being attributed to victims and, by relying on information as the main element of education, may shift responsibility onto the individual. Improvements in the surveillance system should aim to restructure the types of data collected, and facilitate intra-governmental and inter-sector collaboration. The culture of monitoring based on report writing must change to one of surveillance that leads to intervention.
|